AmazonScience/SWE-PolyBench_500

SWE-PolyBench是一个面向代码仓库级别的多语言软件工程基准测试数据集，包含了Python、Java、JavaScript和TypeScript四种语言的实例。数据集分为完整数据集、分层抽样数据集和验证数据集三种类型。每个实例包括实例ID、补丁、仓库信息、基础提交哈希、提示文本、创建日期、测试补丁、问题陈述、测试通过情况等多个字段，用于评估编码代理的性能。

SWE-PolyBench is a multi-language software engineering benchmark dataset for repository level evaluation, including instances in Python, Java, JavaScript, and TypeScript. The dataset consists of three types: the full dataset, a stratified sampled dataset, and a verified dataset. Each instance includes fields such as instance ID, patch, repository information, base commit hash, hint text, creation date, test patch, problem statement, and test pass conditions, which are used to assess the performance of coding agents.